FPO
IEEE

OCEANS’15 Genova

Student Poster Competition, OCEANS’15 MTS/IEEE GENOVA

The poster session and the awards ceremony.

The Student Poster Program has been initiated by Norman Miller in 1989 and became an integral part of the OCEANS conferences in 1991. Since then, more than 700 students have participated in this program. This 36th Student Poster Program of the OCEANS Conferences was held at OCEANS’15 MTS/IEEE Genova, at the Centro Congressi di Genova, from May 18 to May 21. As for the previous Student Poster Competitions, outstanding posters describe the work that the students were presenting and were particularly appreciated by the attendees of the conference. Moreover, the student participants greatly appreciated the opportunity to display, exchange and describe their research work to the community.
     The program was organized by Paola Picco as local coordinator and Philippe Courmontagne, SPC Chair, from IEEE OES. For this 36th edition, 52 abstracts were received, 21 were selected, not without difficulty given the high quality of the received abstracts. The students were from schools in Europe, Asia and the USA. The program was supported by funding from the US Navy Office of Naval Research Global and from AUVSI, which enabled the students to attend the conference.
     The roster of students, their schools (in order of appearance of the Program Book) are:

  • Ashish Agarwal, Indian Institute of Technology, Delhi
  • Mohammadreza Babaee, Technische Universität München
  • Josep Bosch, Universitat de Girona
  • Arnau Carrera Viñas, Universitat de Girona
  • Javier Pérez Soler, Jaume I University of Castellon
  • Ridha Fezzani, ENSTA Bretagne
  • Christopher Gianelli, University of Florida
  • Sheng-Wei Huang, National Taiwan University
  • Guillem Vallicrosa, Universitat de Girona
  • Yann Le Gall, ENSTA Bretagne
  • Graham McIntyre, Dalhousie University
  • Ugo Moreaud, DCNS – Underwater detection Dpt, Acoustic signature R & D
  • Alain Olivier, Department of Information Engineering, University of Padova
  • Benjamin Ollivier, Institut Mines-Télécom Bretagne
  • Albert Palomer, Universitat de Girona
  • Antonio Peñalver Monfort, Jaume I University of Castellon
  • Laurent Picard, Lab-STICC UMR CNRS 6285 ENSTA Bretagne
  • Andrew Stuntz, Fort Lewis College
  • Lingji Xu, Northwestern Polytechnical University
  • Rui Yang, ENSTA Bretagne & Ocean University of China
  • Yang Zhang, Ocean University of China

     The posters were judged by a team organized by IEEE OES. The student award winners were announced during the Gala Dinner at the Palazzo Ducale.
     Dr. Philippe Courmontagne opened the awards ceremony and presented, with Paola Picco, each student with a Certificate of Participation in the OCEANS’15 MTS/IEEE GENOVA. Then, René Garello, IEEE OES President, and Ray Toll, MTS President, presented the third place winner to Mohammadreza Babbaee, from Germany. Next, they presented the second prize to Hugo Moreaud, from France. The first prize, the “Norman Miller’s Prize”, has been presented by Patricia Gruber, Technical Director at Office Naval of Research Global, to Ridha Fezzani, from France, for his poster entitled “Swath bathymetric data fusion – Application to underwater vehicle”. All the students received a round of applause for their accomplishments and participation in the Student Poster Program of Genova.


Ashish Agarwal, Indian Institute of Technology, Delhi
Iterative adaptive approach to DOA estimation with acoustic vector sensors
     Acoustic Vector Sensors are underwater sensors that measure acoustic pressure as well as the acoustic particle velocity to estimate the acoustic intensity. The acoustic intensity, which is a vector quantity, represents the magnitude and direction of the active or propagating part of an acoustic field thus indicating the DOA (Direction of arrival) of a received signal. This paper dwells on an Iterative Adaptive Approach (IAA) for DOA and signal power estimation of underwater acoustic emissions, using a single vector sensor, which is of significance in several passive surveillance applications. The proposed IAA algorithms are robust in resolving partially correlated or coherent sources which is a known phenomenon in shallow water conditions due to multipath. It has been observed that in case of a single vector sensor, the proposed IAA Amplitude and Phase Estimation Scheme (IAA-APES) performs better than a standard Capon beamformer even in the presence of correlated/coherent sources, the performance being measured in terms of Mean Square Angular Error (MSAE). Also another version of IAA, namely the Iterative Adaptive Approach – Maximum Likelihood (IAA-ML) estimation provides better resolution than all the other schemes.

Mohammadreza Babaee, Technische Universität München–The third place winner
Improved Range Estimation and Underwater Image Enhancement Under Turbidity by Opti-Acoustic Stereo Imaging
     Images recorded in turbid waters suffer from various forms of signal degradation due to light absorption, scattering and backscatter. Much of the earlier work to enhance color, contrast and sharpness follow the single-image dehazing approach from the atmospheric imaging literature. Requiring knowledge of both range to scene objects and ambient lighting, various techniques differ in how they estimate the information from various image regions. Moreover, some assumptions are made that hold for most images recorded in air and clear waters, but are often violated in turbid environments, leading to poor results. Alternatively, stereo imaging and polarization have been explored for simultaneous range estimating and image dehazing, however, these can become ineffective with low visibility and (or) weak polarization cue.This work explores a methodology that utilizes the visual cues in multi-modal optical and sonar images, namely, the occluding contours of various scene objects that can be detected and matched more robustly than point features. Calculating the sparse 3-D positions of these contours from opti-acoustic stereo data, we infer a dense range map by exploiting an MRF-based statistical framework, where image intensities and range values serve as observation and hidden variables. Additionally, the opti-acoustic epipolar geometry guides the interference of the MRF by refining neighborhood pixels. The improved performance over other state-of-the-art techniques is demonstrated using images recorded under different turbidity conditions.

Josep Bosch, Universitat de Girona
Creating 360° underwater virtual tours using an omnidirectional camera integrated in an AUV
     The use of omnidirectional cameras underwater is enabling many new and exciting applications in multiple fields. Among these, the creation of virtual tours from omnidirectional image surveys is expected to have a large impact in terms of science and conservation outreach. These surveys can be performed by Remotely Operated Vehicles (ROVs) and Autonomous Underwater Vehicles (AUVs) that can cover large marine areas with precise navigation. Virtual tours are relevant in zones of special interest such as shipwrecks or underwater nature reserves for both scientists and the general public. This paper presents the first results of surveys carried out by an AUV equipped with an omnidirectional underwater camera, and explores the process of automatically creating virtual tours from the most relevant images of the datasets.

Arnau Carrera Viñas, Universitat de Girona
Learning multiple strategies to perform a valve turning with underwater currents using an I-AUV
     Recent efforts in the field of intervention autonomous underwater vehicles (I-AUVs) have started to show promising results in simple manipulation tasks. However, there is still a long way to go to reach the complexity of the tasks carried out by ROV pilots. This paper proposes an intervention framework based on parametric Learning by Demonstration (p-LbD) techniques in order to acquire multiple strategies to perform an autonomous intervention task adapted to different environment conditions. The manipulation skills of a pilot are acquired thought a set of demonstrations done under different environment circumstances, in our case different levels of water current. The proposed algorithm is able to learn these different strategies and depending on the estimated water current, autonomously reproduce a combined strategy to perform the task. The p-LbD algorithm as well as its interplay with the rest of the modules that take part in the proposed framework are described in this paper. We also present results on a free-floating valve turning task, using the Girona 500 I-AUV equipped with a manipulator and a customized end-effector. The obtained results show the feasibility of the p-LbD algorithm to perform autonomous intervention tasks combining the learned strategies depending on the environment conditions.

Javier Pérez Soler, Jaume I University of Castellon
Benchmarking using UWSim, Simurv and ROS: an autonomous free floating dredging intervention case study
     This paper proposes the use of UWSim (an underwater simulator) in combination with Simurv (a kinematic and dynamic library for Underwater Vehicle-Manipulator Systems control algorithms) and ROS (a well-known robotics framework) in order to simulate the dynamics of an Intervention Autonomous Underwater Vehicle and its application to the benchmarking of autonomous control algorithms in the field of archaeology dredging.

 

 

Ridha Fezzani, ENSTA BretagneThe first place winner
Swath bathymetric data fusion – Application to autonomous underwater vehicle
     The autonomous underwater vehicle (AUV) DAURADE platform can acquire bathymetry with two acoustic sensors: a multibeam echo sounder (MBES) and an interferometric sidescan sonar (ISSS). The two sensors (MBES and ISSS) are synchronized and they can simultaneously operate and acquire the bathymetry with different resolutions, geometries and error models. This complementarily allows us to improve the accuracy and the coverage of the collected bathymetric data by fusing both of them. We applied the fusion process on actual data from the two bathymetric sensors of DAURADE (Reson 7125 MBES and Klein 5000 Interferometric); the obtained results are presented and discussed.

 

Christopher Gianelli, University of Florida
Active Sonar Systems in the Presence of Strong Direct Blast
     Active sonar system performance in the presence of strong direct blast, or reverberations from the projected waveform, is considered. The presence of these reverberations may severely limit target detection performance and the applicability of the continuous active sonar (CAS) paradigm. Signal processing approaches to detect objects of interest and estimate relevant target parameters are described for signals subjected to these powerful reverberations. Results obtained by processing data from the TREX’13 experiment are presented and discussed.

 

 

Sheng-Wei Huang, National Taiwan University
Efficient Seafloor Classification and Cable Route Design using an AUV
     This paper aims to an efficient method for submarine cable route design using online seafloor classification from sonar scanlines conducted by an autonomous underwater vehicle (AUV). Currently, the cable route design works are carried out by experienced surveyors and engineers by hand. An online seafloor classification using an AUV with automated route planning method can improve the efficiency for submarine cable construction. Side scan sonar is a common device used for seafloor mapping and obstacles detection. In order to implement online seafloor classification and mapping, an AUV equipped with a side scan sonar is utilized to gather sonar scanlines. Scanlines are analyzed on the fly to classify sea floor using a probabilistic classifier based on Bayes’ theorem and Naïve assumption to distinguish different types of seafloor. Based on the classified seafloor map, a probabilistic roadmap is constructed and an A* algorithm is applied to determine appropriate cable routes on the cable corridor. Seafloor classification, bathymetry, steep slope, angle of alter course, and cable length are the five factors of route design. A result of a cable route survey work between islands was demonstrated. The planned route using the proposed method is close in range to the one recommend by experts.

Guillem Vallicrosa, Universitat de Girona
Simultaneous Mapping and Planning for Autonomous Underwater Vehicles in Unknown Environments
     New potential applications of autonomous underwater vehicles (AUVs) involve operations in unknown and cluttered environments, therefore increasing the vehicle exposure to collisions. To cope with these situations, we use an AUV framework for planning collision-free paths in unknown environments, which adapt and replan the paths according to nearby obstacles perceived during the mission execution using different range sensing sonar. We present simulation and real-world results for the SPARUS-II AUV, a torpedo-shaped vehicle, performing autonomous missions.

 

Yann Le Gall, ENSTA Bretagne
Bayesian source localization with uncertain Green’s function
     The localization of an acoustic source in the oceanic waveguide is a difficult task because the oceanic environment is often poorly known. Uncertainty in the environment results in uncertainty in the source position and poor localization results. Hence, localization methods dealing with environmental uncertainty are required. In this paper, a Bayesian approach to source localization is introduced in order to improve robustness and obtain quantitative measures of localization uncertainty. The Green’s function of the waveguide is considered as an uncertain random variable whose probability density accounts for environmental uncertainty. The uncertain distribution over range and depth is then obtained through the integration of the posterior probability density (PPD) over the Green’s function probability density. An efficient integration technique makes the whole localization process computationally efficient. Some results are presented for a simple uncertain Green’s function model to show the ability of the proposed method to give reliable PPDs.

Graham McIntyre, Dalhousie University
Low-Power Beamforming for Underwater Acoustic Sensing Using a 5-Element Circular Hydrophone Array
     In this paper we present a technique for underwater acoustic beamforming based on soundfield recording that encodes both the temporal and spatial characteristics of a signal. Here, we introduce the basic theory behind soundfield recording and present a first-order beamformer that beamforms the encoded data in a specific direction (i, z) and with a variable polar pattern, p. The appeal of this beamformer being that a 2-dimensional beam can be created using only 4 multiplications and 2 additions. A method for implementing a wideband, planar soundfield recorder using a 5-element array is then discussed. Results from underwater experimental trials using this 5-element array are then presented that compare the measured beam patterns and frequency response of the physical beamformer to the ideal, theoretical beam patterns and frequency response.

Ugo Moreaud, DCNS – Underwater detection Dpt, Acoustic signature R & D – The second place winner
A new way for underwater acoustic signal analysis: The morphological filtering
     This study presents an innovative way for underwater acoustic signal analysis. It is based on multi-directional filters implementation on time-frequency representation, where each filter is designed to enhance a given direction on the time frequency plane. To do so, the proposed technique processes the time-frequency plane by taking into account the actual atom and its neighborhood for each direction, up to a given distance. Thus, such an approach emphases information about signal directional features into the time-frequency plane. Derivation of the technique relies first on the use of a recursive algorithm which estimates noise level. Then, after establishing filters responses for one direction, the design method is extended to all-direction, leading to the morphological filtering, which allows specific morphological feature patterns detection from the time frequency plane. This paper finally presents experimentations on real underwater acoustic recordings to show performances obtained with this technique when objective is SNR enhancement and acoustic signature features preservation.

Alain Olivier, Department of Information Engineering, University of Padova
Modeling the Throughput of 1-persistent CSMA in Underwater Networks
     The aim of this paper is to present a model for the throughput of the 1-persistent CSMA protocol in underwater networks, where the typically large propagation delay with respect to the packet transmission time requires to take into account the spatial distribution of the nodes. Our model is developed based on the analysis carried out in [1] for the non-persistent CSMA protocol. Our results show that the 1-persistent CSMA model developed by Tobagi and Kleinrock is still valid as an approximation, with a few small adjustments, even though it considers an equal propagation delay for all pairs of nodes in the network. The proposed model is validated against simulation results based on the network simulator OMNeT++.

 

Benjamin Ollivier, Institut Mines-Télécom Bretagne
Detection Improvement by Phase Study of the Analytical Cross-Correlation Signal
     In this paper, we propose a test of hypothesis improvement, by phase study of the analytical cross-correlation function in acoustical detection application. Robustness of false alarms probability for the Time Of Arrival (TOA) estimation represents the goal of the proposed method. After signal detection, TOA will be used to localize one receiver, thanks to a grid of transmitters (more than 3), thanks to the knowledge of positions and transmissions times. The presented method is based on a priori information of the researched signal, forming the correlation signal shape (duration and bandwidth). Knowing the auto-correlation peak shape, we will estimate a range with the cross-correlation peak, and deduce if the detected peak corresponds to the researched signal or to a strong noise.

Albert Palomer, Universitat de Girona
Multibeam object segmentation for underwater navigation correction
     Building accurate bathymetries of the seabed has been a focus of study in the last decade. For this purpose seabed point cloud registration has been a focus for some researchers. Some of these registration methods are based on gathering the points of the cloud that contain more information for the registration (i.e. that are flat or smooth, normally being the seabed) and using them as part of ICP-derived methods. For this point picking purpose, we present a segmentation technique that distinguishes between objects (interesting for registration) and ground (smooth and not interesting for registration). The method proposed here uses difference of normals for object’s border detection and a variation of the Density-Based Spatial Clustering of Application with Noise for object clustering. Once the objects boundaries are detected and the points are clustered the rest of the points are classified as object or ground. This classification is done by taking all the points that lie within the object’s border and checking it’s depth compare to its closest border point. The method is evaluated using a multi-beam dataset gathered on the La Lune shipwreck, a site of archaeological interest.

Antonio Peñalver Monfort, Jaume I University of Castellon
Multi-View Underwater 3D Reconstruction using a Stripe Laser Light and an Eye-in-Hand Camera
     Autonomous manipulation in unestructured underwater scenarios is a high challenging skill that has been poorly studied and is becoming more and more important in the last years. One of the main problems regarding the autonomous manipulation, is to find out the characteristics of the object which is going to be manipulated. This paper presents a new approach to obtain an accurate 3D reconstruction of this object. This approach consists in attaching a laser stripe emitter and a camera in the forearm of a robotic arm. Moving the arm, the laser scans the scene where the object is and, at the same time, the camera records the scan. Thanks to the arm and the position of the camera, the scene can be reconstructed from different views and from a position close to the object. The recorded images are processed to obtain the 3D position of the part of the scene projected by the laser. Before the intervention, a process of calibration is needed to calculate the relationship between each part of the system. Furthemore, in order to reduce the time of processing of the images recorded during the scan, an optimization algorithm is presented which consists in discarding, before the processing, the pixels of the image which do not contain relevant information. The approach herein presented and the optimization algorithm are tested using an underwater simulator.

Laurent Picard, Lab-STICC UMR CNRS 6285 ENSTA Bretagne
Potential of the intrinsic dimensionality for characterizing the seabed in the ATR context
     Up to now in mine warfare context, most of Automatic Target Recognition (ATR) processes suffer from environmental effects and therefore try to erase them by image filtering or thresholding before performing shadow and/or echo extraction of the target. In order to design potentially more effective methods, environmental characterization can be investigated. The idea is to incorporate environmental features in the ATR process or to design it according to environmental characteristics such as seafloor homogeneity or complexity. In this way, this paper studies the ability to decompose sidescan sonar images through geometrical structures using the concept of intrinsic dimensionality (iD) and scale-space representations.

 

Andrew Stuntz, Fort Lewis College
Increasing Navigation Accuracy and Localization for Autonomous Gliders to Enable Persistent Autonomy
     To effectively examine ocean processes we must often sample over the duration of long (weeks to months) oscillation patterns. Such sampling requires persistent autonomous underwater vehicles, that have a similarly long deployment duration. Actively actuated (propeller-driven) underwater vehicles have proven effective in multiple sampling scenarios, however they have limited deployment endurance. The emergence of less actuated vehicles, i.e., underwater gliders, has enabled greater energy savings and thus increased endurance. Due to reduced actuation, these vehicles are more susceptible to external forces, e.g., ocean currents, causing them to have poor navigational and localization accuracy underwater. This is exacerbated in coastal regions, where current velocities are the same order of magnitude as the vehicle velocity.
     In this paper, we examine a method of reducing navigation and localization error, not only for navigation, but more so for more accurately reconstructing the path that the glider traversed to contextualize the gathered data, with respect to the science question at hand. We present a set of algorithms for offline processing that accurately localizes the traversed path of an underwater glider over long-term, ocean deployments. The proposed method utilizes terrain-based navigation with only depth, altimeter and compass data compared to local bathymetry maps to provide accurate reconstructions of traversed paths in the ocean.

Lingji Xu, Northwestern Polytechnical University
Robust Sparse Underwater Acoustic Channel Estimation Method Via Projected L1-L2 Optimization
     By exploiting the intrinsic sparse structure of the underwater acoustic channel, we adopt l1–l2 optimization criterion, which incorporates least squares with the penalized l1 minimization to reduce noise effects while inducing channel sparsity. However, the estimate of channel parameters might not be in the feasible region. Therefore, we then define convex sets and utilize convex projection as a supplement to l1–l2 optimization method to further improve the estimation accuracy. In addition, we elaborate matrix multiplications of the l1–l2 solution into vector computation efficiently implemented by FFT. Simulation results show that the proposed method outperforms some conventional channel estimation techniques in low SNR scenarios, and it can resolve multipath effectively even when SNR equals –5 dB.

 

Rui Yang, ENSTA Bretagne & Ocean University of China
Robust Heading Control and its Application to Ciscrea Underwater Vehicle
     Deep inside the ocean, the earth magnetic signal is one of the merely existing information that tells the heading of robots with very good cost efficiency. Therefore, this paper focuses on the AUV (Autonomous Underwater Vehicle) heading control problem using only one magnetic compass as feedback sensor. In this application, we address AUV modeling and control issues simultaneously. Because of quadratic damping factor, underwater vehicle hydrodynamic model is nonlinear. In addition, unmodeled dynamics, parameter variations and environmental disturbances create significant uncertainties between the nominal AUV model and the reality. Finally, sensor noise, signal delay as well as unmeasured states also affect the stability and control performance of AUV motions. In order to handle these issues with improved AUV observation quality and navigation ability, we propose a CFD (Computational Fluid Dynamics) model based H1 robust control scheme. Without loss of generality, the robust heading controller was implemented and validated in the sea on low-mass and complex-shaped Ciscrea AUV. Simulation and sea experimental results of both PID (Proportional Integral Derivative) and robust heading controller are analysed.

Yang Zhang, Ocean University of China
Seafloor Image Compression Based on Hybrid Wavelets and Directional Filter Banks
     We present an efficient compression method based on the hybrid wavelets and directional filter banks (HWD), to achieve high compression efficiency while keeping visually pleasing reconstruction quality for underwater images. According to the characteristics of underwater images and the human vision system (HVS), an improved just noticeable distortion (JND) model is initially employed to adaptively remove the visual redundancy of underwater images in the HWD domain. The low-frequency coefficients are then quantized in the fixed length and encoded losslessly. The high-frequency coefficients are quantized by variable precision and fixed length method, and are coded by the difference reduction algorithm based on HWD trees. The experimental results show that the proposed compression algorithm provides both high coding efficiency and satisfactory reconstruction quality, which is highly desired for the transmission of underwater images at very low-bit rates.